NARSTO COMPENDIUM

R&D METHODS FOR PAMS/VOCS OZONE PRECURSORS

Differential Optical Absorption Spectrometry (DOAS)

A.    Basic Principles

Spectrometry is a chemical analysis technique that makes use of the features of electromagnetic radiation, including light. This radiation proceeds through space in a form which has a measurable wavelength, and a measurable energy or intensity. When the radiation passes through a substance such as the atmosphere it is modified and distorted by the molecules of the many chemical components which are mixed together in the substance.
Differential optical absorption spectrometry (DOAS) measures the absorption through an atmospheric path (typically 0.5 to 1.5 km) of two closely spaced wavelengths of light from an artificial source. One wavelength is chosen to match an absorption line (wavelength) of the compound of interest, and the other is close to but off that line, and is used to account for atmospheric effects. (EPA/600/P-93/004aF, p.3-98).

The term "absorption line" is used to mean a wavelength that a given atom or molecule absorbs more than it does other wavelengths. Thus, if a compound of interest were ozone, a wavelength would be identified which ozone - and ozone especially - absorbs more than other wavelengths.

A light beam containing this particular wavelength would be directed at a segment of the atmosphere, and the amount of the wavelength absorbed would be measured. This result would indicate how much ozone was present in that segment. A separate wavelength, varying slightly from the first, would be directed through the same segment, and its absorption would be compared with the first wavelength results to determine inadvertent absorption probably not due to ozone. This technique provides a comparison absorption level to improve accuracy in measuring the amount of radiation absorbed by the target molecule.

B. Range:

Span of species concentration measured by various DOAS instruments that were found in the literature are, for example:
 
SPECIES LOW (ppbv) HIGH (ppbv)   REFERENCE
p-xylene  2 280  Axelsson et al. (1995),  p1259
ethylbenzene 4 450 Axelsson et al. (1995),  p1259
1,2,3-trimethylbenzene 2  2700  Axelsson et al. (1995),  p1259
NO2 9  135  Biermann et al. (1988), p1551
NO3 0.02 70 Biermann et al. (1988), p1551
Ozone (O3)  10 22 Stevens et al. (1993), p233
SO2 10  50 Stevens et al. (1993), p233

C. Minimum Detection Level:

Minimum Detection Level is defined as the lowest concentration of a species that DOAS devices can distinguish from background noise.
 
SPECIES Minimum Detection Limit 
(ppbv)		  REFERENCE
p-xylene 0.3  Axelsson et al. (1995), p1258
ethylbenzene  2 Axelsson et al. (1995), p1258
1,2,3-trimethylbenzene 6 Axelsson et al. (1995), p1258
NO2 4 Biermann et al. (1988), p1551
NO3 0.02 Biermann et al. (1988), p1551
Ozone (O3)  3 Stevens et al. (1993), p234
SO2 10 Stevens et al. (1993), p234

D.  Operating Temperature:

An ambient air DOAS system was operated at an altitude of 15 meters above ground level in Lubbock, Texas (Vecera and Dasgupta, 1991). The average monthly temperature for April at Lubbock is about 18oC (Rudloff, 1981). In September and October, a DOAS light transmitter and receiver were operated on the roofs of EPA buildings, 20 meters above ground level in Research Triangle Park, North Carolina (Stevens et al., 1993). The average monthly temperature is estimated at 23oC and 18oC for those two months (Rudloff, 1981). In Jonkoping, Sweden, twelve months of continous ozone monitoring were performed using a 750 meter retroreflector path, giving a total path length of 1500 meters (Axelsson et al., 1990). The average monthly temperatures range between -3oC to 16oC (Rudloff, 1981). E. Known Interferences: a. In open path DOAS applications, interference by miscellaneous atmospheric constituents is likely to be a problem. For example, in detecting ozone, oxygen is likely to interfere in the region below 270 nm. Sulphur dioxide has a strong absorption effect in the vicinity of 300 nm, and also presents interference, although weak, at 283 nm (Axelsson et al., 1990). Another example of species interference would be NO and NO2 interfering with detection of HONO (Vecera and Dasgupta, 1991).   b. Heavy rain and fog, and even high humidity, have been found to interfere with the xenon lamp beam propagation, and to make absorption measurements impossible. The use of heaters on the lenses did not entirely solve this problem. However, it only becomes serious in very extreme conditions (Stevens et al., 1993). c. In long path uses of UV light beams, atmospheric turbulence, such as that from thermal-induced effects, can distort reflections. One approach to solving this problem has been to use an extensive array of photodiode type receivers, but these arrays must all be calibrated, and can suffer from differences in surface coatings. Turbulence can be approached with other strategies, such as slotted disks, and new systems can be developed with these strategies (Edner et al., 1993).  With advent of solid-state CCD and/or CMOS detectors, it is possible that a more robust detector can be built.

F.  Notes of Interest:

a. In selecting any long path pollution monitoring strategy, DOAS should be compared with DIAL (Differential Absorption Lidar), a technique using pulsed lasers, either with wavelength switching, or with two separate wavelength laser sources. Although UV DOAS has problems with atmospheric interference, as noted above, it has not been abandoned in favor of the laser approach.   b. The DOAS approach has probably benefitted by the relatively simple applications which investigators have favored. In the meantime, DIAL and LIF ( Laser induced fluorescence) may well have suffered from attempts to develop airborne and other mobile equipment. All of these techniques are still in the developmental stage, but the DOAS should continue to enjoy the advantage of an established beam source, the arc lamp, while laser sources include the tunable diode and the YAG, and may require the use of pumping and other ancillary technology. G.  References: